Journal of Crystal Growth 240 (2002) 230–235

A laboratory investigation of cyanobacterial extracellular

polymeric secretions (EPS) in influencing CaCO3
polymorphism
T. Kawaguchi*, A.W. Decho
Department of Environmental Health Sciences, Norman J. Arnold School of Public Health, University of South Carolina,
Columbia, SC 29208, USA

Accepted 31 January 2002

Communicated by N. Chayen

Abstract

Bahamian stromatolites are well-laminated structures, consisting of lithified layers alternating between unlithified
layers containing fine-grained carbonate ooids. The lithified layers consist of abundant aragonite needles embedded
within a matrix of extracellular polymeric secretions (EPS) by cyanobacteria, Schizothrix sp. Laboratory investigations
were conducted using EPS extracted from natural stromatolites and laboratory isolates of Schizothrix sp., to chemically
characterize EPS, and determine in vitro how EPS may influence CaCO3 polymorphism. EPS mainly consisted of acidic
polysaccharides and proteins. Biochemical analyses indicated that contents of uronic acids and carbohydrates in EPS
from lithified layers decreased when compared with unlithified layer EPS, while the protein content remained relatively
constant. CaCO3 nucleation experiments demonstrated that EPS from the lithified layer, induced aragonite crystal
formation in vitro, as confirmed by scanning electron microscopy and Fourier transform infrared (FT-IR)
spectroscopy. In contrast, EPS from the unlithified layer or laboratory-cultured Schizothrix sp. induced calcite crystal
formation. These laboratory results suggest the possibility that the biochemical composition, specifically small proteins,
of EPS influences the resulting mineralogy of CaCO3. r 2002 Elsevier Science B.V. All rights reserved.

Keywords: A1. Nucleation; A2. Growth from solutions; A2. Seed crystals; B1. Biological macromolecules; B1. Calcium
compounds; B1. Polymers

1. Introduction examples of stromatolites presently forming in

Marine stromatolites, found on the margins of
Exuma Sound, Bahamas are the only known
*Corresponding author. Tel.: +1-803-777-4144; fax: +1-
803-777-3391.
E-mail address: tkawaguchi@sophe.sph.sc.edu
(T. Kawaguchi).
open marine environment of normal seawater
salinity [1]. The surfaces of Highborne Cay
stromatolites are well laminated, consist of alter-
nating lithified and unlithified layers. The lithified
layers typically contain an abundance of amor-
phous extracellular polymeric secretions (EPS), a
metabolically diverse community of heterotrophic
microorganisms [2,3] and aragonite needles. Near

0022-0248/02/$ - see front matter r 2002 Elsevier Science B.V. All rights reserved.
PII: S 0 0 2 2 - 0 2 4 8 ( 0 2 ) 0 0 9 1 8 - 1
 T. Kawaguchi, A.W. Decho / Journal of Crystal Growth 240 (2002) 230–235
the surfaces of stromatolites, unlithified layers are
typically laden with copious EPS produced by
cyanobacteria [2]. The cyanobacterium, Schizo-
thrix sp., is a dominant species that secretes
EPS in this system [1]. Aragonite precipitates,
approximately 1 mm in length, form spherical
aggregates 2–5 mm in diameter and are embedded
in the EPS matrix [1]. Bacteria are abundant and
commonly observed at the edges of the aragonite
spherules. Unlithified layers alternate between
lithified layers, and are sedimentary structures
composed mainly of fine-grained, well-sorted
carbonate ooids, i.e., sand, approximately 125–
250 mm. The EPS matrix is a key structuring
component of microbial biofilms [4,5]. Many EPS
are widely produced by microbes for attachment
and protection, and serve important roles in
providing nucleation sites and facilitating sediment
trapping [6]. EPS predominantly contain acidic
polysaccharides, mainly uronic acids, and proteins
[4,7].
It has been suggested that the negative charges
of uronic acids play important roles in the various
functions mentioned above [4]. Aragonite
precipitation in marine stromatolites is considered
to be a biologically induced mineralization
and a by-product of several microbial processes
[1]. The mechanism of induction for aragonite
precipitation in marine stromatolites may be
related to specific interactions of microbial
photosynthesis, respiration and sulfate reduction.
Magnesium is known to induce aragonite forma-
tion in seawater and in vitro at ratios of Mg/
Ca>4, but has never been demonstrated in vivo
[8]. Therefore, factors influencing CaCO3 poly-
morphism in marine stromatolite are not clearly
understood.
In the present study we have focused on
aragonite precipitation by EPS under in vitro
conditions. In this study, EPS isolated from both
‘‘lithified’’ and ‘‘unlithified’’ layers of natural
marine stromatolites, and from laboratory cul-
tured stromatolite forming cyanobacterium, Schi-
zothrix sp. mats were used to examine specific
influences on aragonite precipitation. CaCO3
nucleation experiments were conducted to deter-
mine if EPS influences aragonite precipitation in
vitro.
231
2. Experimental procedure
2.1. EPS isolation
Samples of intertidal marine stromatolites were
collected from Highborne Cay, Bahamas
(761490 W, 241430 N). Samples were returned to
the laboratory, and EPS were scraped off from
both lithified and unlithified layers of the samples
using a sharp, pointed scalpel under a dissecting
microscope. Only the upper 2 mm of mat were
used to isolate natural mat layers for EPS. The
‘‘lithified’’ and ‘‘unlithified’’ layers were carefully
separated under a dissecting microscope. The
isolated layers, consisting of cells, sediment, EPS,
and precipitate fragments were placed in 1.5 ml
Eppendorf tubes and suspended in 0.05 mM
EDTA solution, heated at 401C, and stirred for
30 min.
The cyanobacteria, Schizothrix sp., was isolated
from marine stromatolites at Highborne Cay,
Bahamas. This species was grown in CHU-10
medium [9] consisting of 0.004 M Na2SiO3  9H2O,
0.006 M Ca(NO3)2  4H2O, 0.014 M K2HPO4,
0.025 M MgSO4  7H2O, 0.05 M Na2CO3, 0.012 M
Fe-EDTA, 3.7  108 M B12, 4  107 M biotin,
and 5.9  107 M thiamine in seawater of 32 ppt
salinity on a light:dark cycle of 12:12 h, at
approximately 100 mEinsteins. Bahamian sedi-
ment, consisting of well-sorted CaCO3 ooids
(mean grain size 100–250 mm), was collected from
the study site, sterilized, cleaned in sodium
hypochlorite and rinsed thoroughly in distilled
H2O. The sediment was added to culture flasks to
a depth approximately 0.5 cm as a substratum for
growth. The cultures were grown for several weeks
until a firm mat of cyanobacteria was present on
the surface of the sediment.
For isolation of EPS from laboratory cultured
Schizothrix sp., culture medium was discarded,
then each of the cultures was suspended, stirred,
and heated to 401C for at least 30 min to strip EPS
from the cyanobacteria. Then, the separated
suspensions from both natural and laboratory
samples were centrifuged at 12000 rpm for 15 min
to shear remaining EPS from cells. To remove
small-molecular weight (MW) components, the
supernatant was dialyzed (MW cutoff 14,000 Da)
232 T. Kawaguchi, A.W. Decho / Journal of Crystal Growth 240 (2002) 230–235
in de-ionized water with constant stirring for 48 h.
The dialyzed solution was lyophilized and stored
at 701C.
2.2. Chemical analysis of EPS
The phenol-sulfuric acid spectrophotometric
method was used to measure carbohydrate content
as hexamine [10]. Uronic acid content was
measured using carbazole in 80% sulfuric acid
with borate ions added [11]. Protein content was
measured according to Bradford’s method using
BioRad Protein Assay Kit [12]. All samples were
measured in triplicate.
2.3. CaCO3 nucleation experiment
EPS from lithified and unlithified layers, and
laboratory-cultured Schizothrix sp. were used for
this experiment. EPS were fixed using 4% for-
maldehyde on to solid agarose beads [13]. Then,
the EPS attached to agarose beads were placed in
solutions containing 10 mM CaCl2, 5 mM NaH-
CO3, and 30 mM MgCl2, and incubated for 5 days
[14,15] in scintillation vials. Agarose beads,
containing no EPS coating, were used as controls.
At the end of the experiment, agarose beads were
filtered on a 0.45 mm membrane filter. The filter
was washed with dilute NH4(OH)(pH 8), air-dried,
mounted on a specimen holder, sputtered with
gold, and examined with an Hitachi Delta 2500
scanning electron microscope (SEM). The type of
crystal formed on agarose beads was investigated
using a Kevex energy dispersive X-ray spectro-
meter (EDAX) attached to the SEM. Attenuated
total reflectance (ATR) Fourier transform infrared
(FT-IR) spectra were obtained using NEXUS 670
FT-IR (Thermo-Nicolet, Inc.), equipped with a
germanium internal reflection element (Spectral-
Tech. Inc.).
3. Results
Table 1 shows the contents of carbohydrate,
uronic acids and protein in EPS both from lithified
and unlithified layers, and laboratory-produced
EPS. Carbohydrate and uronic acid content
Table 1
Contents (w/w) of carbohydrate, uronic acids and protein in
EPS from lithified and unlithified layers and laboratory
cultured Schizothrix sp. biomat in marine stromatolites,
Hayborne Cay, Bahamas
Carbohydrate Uronic acids Protein
(%) (%) (%)
Lithified layer 0.9 4.3 2.5
Unlithified layer 17.24 15.27 2.8
Schizothrix sp. 22.4 19.34 1.9
(17.24% and 15.27%, respectively) in the unlithi-
fied layer EPS were higher than those (0.9% and
4.3%, respectively) in lithified layer EPS. How-
ever, carbohydrate and uronic acids contents
(22.4% and 19.34%) in laboratory-produced EPS
were similar to those EPS collected from the
unlithified layers. Protein contents from lithified
layers, unlithified layers and laboratory produced
EPS were 2.5%, 2.8% and 1.9%, respectively.
After 5 days of incubation, CaCO3 minerals
were present on the surface of EPS-coated agarose
beads (Fig. 1). In controls, lacking EPS, CaCO3
minerals were not observed (data not shown). The
presence of Ca in the precipitates was confirmed
using EDAX analyses. Crystals formed on the
lithified layer EPS-coated agarose beads were
hexagonal, and spindle-shaped (Fig. 1C). FT-IR
spectra revealed peaks splitting near 700 cm1 (711
and 699 cm1), that are characteristic of the
aragonite structure (Fig. 2C) [16] and confirmed
that the CaCO3 crystals formed on the agarose
beads were aragonite. In contrast, the shape of
crystals formed on the unlithified layer (Fig. 1A)
and laboratory cultured Schizothrix sp. EPS
(Fig. 1B) were rhombohedral and FI-IR analysis
confirmed that those crystals formed on agarose
beads were calcite (Fig. 2A and B).
4. Discussion
Aragonite needles are commonly found within
the EPS matrix in the surface layers of stromato-
lites [1]. Similar aragonite needle formation, found
in Halimeda cylindracea, is known to be associated
 T. Kawaguchi, A.W. Decho / Journal of Crystal Growth 240 (2002) 230–235 233

Fig. 1. CaCO3 crystals formed on agarose beads coated with different EPS. (A) Unlithified layer EPS. Scale bars: 500 mm (A-1), 0.5 mm
(A-2), (B) Schizothrix sp. EPS. Scale bars: 500 mm (B-1), 0.5 mm (B-2). (C) Lithified layer EPS. Scale bars: 500 mm (C-1), 0.5 mm (C-2).

Fig. 2. Fourier transformed infrared (FT-IR) spectrum of calcium carbonate crystals formed on agarose beads which coated with
unlithified layer EPS: calcite (A), Schizothrix sp. EPS: calcite (B) and lithified layer EPS: aragonite (C).
234 T. Kawaguchi, A.W. Decho / Journal of Crystal Growth 240 (2002) 230–235
with organic material [17]. Nakahara and Beve-
lander [18] found that organics were coating
Halimeda incrassata aragonite needles. Biochem-
ical analysis of organic materials found in Hali-
meda [19] showed that they are largely
polysaccharides with some proteins which are very
similar to the composition of EPS found in marine
stromatolite. However, the roles of those organics
in nucleation and in determining the polymorph-
ism of CaCO3 are not well understood.
In this study, FT-IR spectra confirmed that the
crystals that formed on lithified layer EPS coated
agarose beads were aragonite, but crystals formed
on unlithified layer and laboratory cultured
Schizothrix sp. EPS coated agarose beads were
calcite. Biochemical analysis showed that contents
of uronic acids and carbohydrates in EPS from
lithified layers were decreased when compared to
unlithified layer EPS, but protein content re-
mained relatively constant. Previously, we have
shown [20], using pH drift assays with chemically
modified EPS, that protein fractions are more
important in CaCO3 precipitation than uronic
acids in the lithified layer. In contrast, both protein
and carboxylated polysaccharides are involved in
the inhibition of CaCO3 precipitation in unlithified
layer [20]. These results suggest that as microbial
degradation of EPS proceeds, specific protein
fractions in EPS may become increasingly ex-
posed.
The results of the present study showed that
EPS, which differed in composition, induced the
precipitation of different CaCO3 crystal types in
vitro. Previously, it was proposed that organic
components in biominerals may determine the
type of CaCO3 [21]. Wada et al. [22] reported that
acidic polysaccharides isolated from calcareous
algae induced Mg-rich calcite precipitation in a
double-diffusion experiment. Acid polysaccharides
inhibited the growth of aragonite crystals by
sorption onto the surfaces of the crystals [22].
Also, Albeck et al. [23] suggested that the
polysaccharides contained in glycoproteins may
modulate calcite crystal growth in vitro. Since
unlithified layer and laboratory-cultured Schizo-
thrix sp. EPS contained greater relative amounts
(Table 1) of acidic polysaccharides than the
lithified layer EPS, calcite, rather than aragonite
formation was selectively induced by those EPS in
the nucleation experiment.
It has been demonstrated that acidic proteins
isolated from aragonitic biominerals induce ara-
gonite formation [24,25]. Levi et al. [26] reported
that the specific amino acid sequence (Asp–Leu)n
in protein isolated from aragonitic biominerals
was capable of inducing aragonite formation. In
the present study, amino acid compositions of
acidic proteins in EPS from marine stromatolites
[20] showed similarity, being specifically enriched
in aspartic acid and glutamic acid, when compared
with key amino acids (aspartic acid and glutamic
acid) found in the organic matrices associated with
other biominerals [27–30]. Changes in stereoche-
mical structure, resulting from attachment of
acidic proteins to a solid surface (e.g. chitin), have
been suggested to be an important requirement for
CaCO3 nucleation and CaCO3 polymorphism
[26,31–33]. Although our results are preliminary,
the data suggest that biochemical composition in
EPS influence the precipitation, composition, and
CaCO3 polymorphism in the lithified layers of
marine stromatolites.
Acknowledgements
We thank the R.V. Calanus (Univ. Miami) for
efficiency and comfort during Bahamian stroma-
tolite sampling; Dr. R.Z. LeGeros, Calcium
Phosphate Research Laboratory, David B. Kriser
Dental Center, College of Dentistry, New York
University for the consultation of identifying the
type of calcium carbonate crystal; Dr. N. Watabe
for carefully reviewing this manuscript. This work
was supported by grants from the National
Science Foundation (OCE 96-17738) and the
Office of Naval Research (NOOO14-97-10024).
This represents RIBS (Research Initiative on
Bahamian Stromatolites) Contribution No. 20.
References
[1] R.P. Reid, P.T. Visscher, A.W. Decho, J.F. Stolz, B.M.
Bebout, C. Dupraz, I.G. Macintyre, J. Pinckney, H.W.
 T. Kawaguchi, A.W. Decho / Journal of Crystal Growth 240 (2002) 230–235
Paerl, L. Prufert-Bebout, T.F. Steppe, D.J. Des Marais,
Nature (London) 406 (2000) 989.
[2] P.T. Visscher, R.P. Reid, B.M. Bebout, S.E. Hoeft, I.G.
Macintyre, J.A. Thompson, Am. Mineral. 83 (1998) 1482.
[3] P.T. Visscher, R.P. Reid, B.M. Bebout, Geologica 28
(2000) 919.
[4] A.W. Decho, Oceanogr. Mar. Biol.: Annu. Rev. 28 (1990)
73.
[5] J.W. Costerton, Z. Lewandowski, D. DeBeer, D. Caldwell,
D. Korber, G. James, Annu. Rev. Microbiol. 49 (1995)
711.
[6] R. Riding, Sedimentology 47 (2000) 179.
[7] J. Trichet, C. DeFarge, Inst. Oceanogr. Bull. (Monaco) 14
(1994) 203.
[8] Y. Kitano, Bull. Chem. Soc. Japan 35 (1962) 1973.
[9] R. Rippka, J. Deruelles, J.B. Waterbury, J. Herdman,
R.Y. Stanier, J. Gen. Microbiol. 111 (1979) 1.
[10] M. Dubois, K.A. Gilles, J.K. Hamilton, P.A. Rebers, F.
Smith, Anal. Chem. 28 (1953) 350.
[11] K.A. Taylor, J.G. Buchanan-Smith, Anal. Biochem. 201
(1992) 190.
[12] M.M. Bradford, Anal. Chem. 72 (1976) 248.
[13] M.E. Gunthorpe, C.S. Sikes, A.P. Wheeler, Biol. Bull. 179
(1990) 191.
[14] M.A. Crenshaw, in: S. Suga, H. Nakahara (Eds.),
Mechanisms and Phylogeny of Mineralization in Biologi-
cal Systems, Springer, Tokyo, 1991, p. 101.
[15] E.M. Greenfield, D.C. Wilson, M.A. Crenshaw, Am. Zool.
24 (1984) 925.
235
[16] J. Balmain, B. Hannoyer, E. Lopez, J. Mater. Res. 48
(1999) 749.
[17] M.A. Borowitzka, Int. Rev. Cytol. 74 (1982) 127.
[18] H. Nakahara, G. Bevelander, Jpn. J. Phycol. 26 (1978) 9.
[19] M.A. Borowitzka, A.W.D. Larkum, J. Phycol. 13 (1979) 6.
[20] T. Kawaguchi, A.W. Decho, Prep. Biochem. Biotech, in
press.
[21] N. Watabe, K.M. Wilbur, Nature (London) 188 (1960)
334.
[22] N. Wada, M. Okazaki, S. Tachibana, J. Crystal Growth
132 (1993) 115.
[23] S. Albeck, S. Weiner, L. Addadi, Chem. Eur. J. 2 (1996)
278.
[24] Q.L. Feng, G. Pu, Y. Pei, F.Z. Cui, H.D. Li, T.N. Kim, J.
Crystal Growth 216 (2000) 459.
[25] G.S. Zhang, X.D. Xie, J. Inorg. Mater. 15 (4) (2000) 765.
[26] Y. Levi, S. Albeck, A. Brack, S. Weiner, L. Addadi, Chem.
Eur. J. 4 (3) (1998) 389.
[27] M. Akiyama, Proc. Jpn. Acad. Sci. 42 (7) (1966) 800.
[28] T. Kawaguchi, N. Watabe, J. Exp. Mar. Biol. Ecol. 170
(1993) 11.
[29] R.J. Kingsley, N. Watabe, Comp. Biochem. Physiol. 76-B
(1983) 443.
[30] R.M. Mitterer, Bull. Mar. Sci. 28 (1978) 173.
[31] S. Weiner, Cal. Tissue Int. 29 (1979) 163.
[32] S. Weiner, J. Erez, J. Foraminifer Res. 14 (1984) 206.
[33] A. Sugawara, T. Kato, Chem. Commun. 6 (2000) 487.